1. Field of Invention
The present invention relates to an improved bi-directional laser communications link between two optical transceiver platforms employing improved wavefront sensing (WFS) compensation mechanisms as a way of and a means for producing the highest possible resolution at the focal plane or maximizing the power into a communications fiber, and eliminating the physical misalignment between the communications fiber and the wavefront sensor, so as to improve communication link performance.
2. Brief Description of the State of the Art
There is great interest in super broad-band free-space optical (FSO) laser communication systems as they are capable of securely communicating information at high data rates in point-to-point and multi-point-to multi-point communication networks.
In military applications, such laser communication systems and networks offer a level of superiority and security over radio-frequency (RF) based communication system which have relatively limited band-widths, and thus data transfer rates, as well as being susceptible to RF-based jamming techniques intended to interfere and disrupt the performance of such systems. In commercial applications, such laser communication systems can be rapidly installed in point-to-point and multi-point-to-multi-point configurations (using buildings and towers as support structures for such laser communication platforms) at a significantly reduced expense in comparison with micro-wave-based satellite communication systems. Examples of such FSO laser communications systems are disclosed in U.S. Pat. Nos. 6,657,783; 6,643,467; 6,348,986; 6,347,001; 6,314,163; 6,285,481; 6,286,944; 6,181,450; 6,151,340; 6,122,084; 5,844,705; 5,786,923; 5,786,923; 5,754,323; 5,710,652; 5,606,444; and WIPO Publication No. WO 03/003618, each said prior art reference being incorporated herein by reference.
However, FSO laser communication systems are not without challenges and problems and challenges. In particular, the free-space optical communication links in FSO laser communications are effected by atmospheric conditions such as turbulence and the like which works to cause aberrations in the spatial phase of the wavefront of the modulated carrier laser beams as such carrier laser beams are transmitted between terminals in such communication systems. Depending on the distances between the laser communication terminals, such spatial phase aberrations evolve into spatial intensity aberrations in the laser beam received at the entrance pupil of the receiver module of such terminals, degrading the bit error rate (BER) of such communication systems.
Studies have been conducted by Adaptive Optics Associates, Inc. to quantify these wavefront errors over the range of possible link geometries. The uplink and downlink have been considered separately because the atmosphere affects the propagation differently depending on its direction. The density of the atmosphere declines rapidly with altitude and thus most of the wavefront errors are generated close to the aircraft. Integration of the analytical model of the atmosphere proposed by Fried (JOSA, 56,1380, 1966) shows that, for the altitudes of interest in this study, the effective distance of the turbulence from the aircraft is about 3 km.
For the light propagating up to the aircraft the wavefront errors picked up close to the aircraft evolve along the long path to the satellite so that the beam that reaches the satellite has both phase and intensity errors. This behavior is described by the eikonal equation (Born and Wolff, Principles of Optics, p. 112), which shows that the change in intensity between to planes perpendicular to the propagation direction is proportional to the product of the curvature of the wavefront and the propagation distance. This effect is the source of scintillation in the beam.
For light coming down to the aircraft, the situation is very different. The wavefront has only 3 km to evolve before reaching the receiver and thus is almost purely a phase error without any scintillation. Therefore, the airborne to space borne direction as being the most problematic direction in this link. This fundamental difference leads to very different requirements for the satellite and aircraft transceiver compensation systems.
Modeling of Atmospheric Turbulence
To assess the impact of atmospheric turbulence on the communications link, a model of the atmosphere has been developed. This model is based upon one of the standard models (e.g. Clear 1 model developed for the ABL program) supplemented with field data from the ABL-EX experiment, data from astronomical observatories (Masciadri, et al. Astron. Astrophys. Suppl. Ser. 137,185-202, 1999.), and the AFRL 3D modeling project. Of particular importance in this study is the modeling of the turbulent layer associated with the tropopause, determination of the range of altitude for the tropopause, and modeling of the effects of a jet stream. This is because, unlike an observatory that can be sited at a location known to have very low turbulence, the communications link must operate even if the aircraft is below the tropopause and a strong jet stream is traversed at some zenith angles.
Using the atmospheric, turbulence profiles along various link paths have been determined. The important parameters are the altitude of the aircraft, the velocity of the aircraft, and the altitude and orbital path of the satellite. The aircraft altitude determines the starting point of the link path. The aircraft velocity sets the rate at which the beam translates through the atmosphere. The satellite altitude and path determine the apparent slew rate of the beam. FIG. 1 describes these platform parameters. Once a set of link paths is defined from these parameters, the atmospheric model has been used to create a series of phase screens at different altitudes. These are inputs to the AOA Wave Optic Propagation Code that is used to calculate the effects on the laser beam. Such modeling has been performed for both the uplink and downlink directions. One additional input to the Wave Optic Propagation Code is the aero-optical effect of the flow field around the window of the transceiver on the aircraft. This is modeled as a time varying phase screen. The results of this simulation will be a time sequence of phase and intensity maps for the entrance apertures of the two receivers.
In FIG. 1B a typical phase map calculated for a horizontal communications link with strong turbulence. FIG. 1C shows a time sequence of detected laser beam spot intensity distributions calculated using a simple model atmosphere for propagation to a satellite at 1000 km altitude from an aircraft at 35,000 feet. The bar in each frame represents 10 m at the satellite. The second through fourth frames in this time sequence show a dark (rather than bright) speckle of several meters extent. Such an event would lead to a deep fade of the link signal for any receiver aperture of less than a few meters diameter.
The atmospheric model described above is used to examine other aspects of the laser communication link compensation problem. For example, because of the long path to the satellite, the light travel time to the satellite becomes significant. The result is that the transmitted beam must be pointed ahead of the apparent location of the satellite, just as a hunter must point his gun ahead of a flying duck to successfully hit it. This point-ahead angle is typically tens of microradians. Difficulties arise when this point-ahead angle becomes larger than the isoplanatic angle of the atmospheric turbulence. This is the angle over which the aberration is correlated. If the point-ahead angle is larger than the isoplanatic angle, the wavefront error measured from a source co-located with the satellite does not correctly represent the aberrations that will be experienced along the path of the transmitted beam. Also, as wavefront tilt is the strongest atmospheric aberration, this aniosplanatic effect can lead to serious pointing errors in the transmitted laser beam if the employed laser beam tracking algorithm uses only the satellite apparent position to estimate the error. A great body of research exists on this effect, both in the astronomical literature and from HEL DEW projects. If tilt anisoplanatism appears to be a significant problem for a given application, then more advanced tracking algorithms need to be drawn from this prior research and tested in the simulation.
Using the above-described techniques, some initial modeling has been performed to verify that atmospheric turbulence is a significant issue for aircraft to space optical communications links. A model atmosphere was constructed based on the analytical model of Fried (JOSA, 56,1380, 1966). The Fried model was modified to include the observed “bump” in turbulence in the region of the tropopause and the strongly layered structure of turbulence. FIG. 1D shows a typical Cn2 profile generated by this model. This model was used to estimate the integrated turbulence along a vertical path from aircraft at various altitudes. The results for one realization of turbulence are shown in FIG. 1E. From these integrated turbulence estimates pertinent parameters for the design of a compensation system were derived. These included the atmospheric coherence length (Fried's ro parameter) and the coherence angle, known as the isoplanatic angle. These calculations were made using a wavelength of 1550 nm.
FIGS. 1F1 and 1F2 show some typical results derived from modeling this phenomenon. In particular, FIG. 1F1 shows that for altitudes below the tropopause (about 45,000′ in this case) the coherence length is comparable to, or smaller than potential transceiver aperture sizes. This means that significant atmospheric aberrations will occur at these altitudes. FIG. 1F2 shows that at those same altitudes, the isoplanatic angle is in the range of tens of microradians. Since this is comparable to the point-ahead angle required for a link to a LEO satellite, the issue of tilt anisoplanatism must be treated carefully in any particular application.
Wave-Optic Propagation Simulations
To further assess the impact of this level of turbulence on a free space optical communication link, some simple wave-optic propagation simulations were performed. Using the integrated Cn2 value, a single phase screen was generated for an aircraft altitude of 35,000 feet. FIG. 1G shows a pseudo-color representation of such a phase screen. The full extent of the screen is 1 m×1 m. Subsections of this phase screen were extracted along a path to simulate the motion of the aircraft past the turbulence. In this simple model, the turbulence itself is frozen in time. The extracted phase screen was then applied to a uniformly illuminated circular pupil of 150 mm diameter and the resulting wavefront propagated to a range of 1000 km.
While the use of a single phase screen does not model the real three-dimensional distribution of the turbulence, because the atmospheric density falls off fairly rapidly with height, it is adequate for obtaining a rough idea of the effects of turbulence on the beam pattern at the satellite. FIG. 1H shows two examples of the beam profile at the satellite compared with the diffraction limited profile (left image). Each image covers about an area 100 m by 100 m. In the case of the center beam profile, little power is lost due to the aberration. However, the laser beam profile on the right shows that significant fade events can occur as the receiver aperture could easily lie within one of the dark (speckle) regions of the laser beam spot image.
A sequence of 576 beam profiles was generated during wave optic propagation simulation. For a platform velocity of 200 m/s this sequence represents about 0.1 s of flight time. The temporal variation of the received power was calculated for this sequence and is shown in FIG. 1I. There is one fade event of about 10 db in this data set. Thus, in any link compensation system, it will be necessary to determine how well it suppresses these occasional fade events in addition to the expected improvement in link power due to the improved beam pattern.
The effect on laser beam pointing has also been examined. For each profile, the location of the centroid of the light was calculated. The rms variation in pointing direction was found to be 12.6 μrad with a peak deviation of 28.7 μrad. This rms variation in pointing is slightly larger than the diffraction limited beam radius and thus must be corrected to maintain an adequate communication link signal.
To determine the parameters for a compensation system, the spatial frequency content of the phase aberrations must be examined. Each of the extracted phase screens is decomposed into Zernike polynomials and the temporal intensity fluctuations of the Zernike coefficients determined. FIG. 1J displays the rms variation in each of the lowest 15 Zernike terms. Notably, the bulk of the aberration is contained in the first three terms, the two tilt terms and defocus. That is typical of all atmospheric aberrations. Correction of just these three terms, in transmission and reception channels of the system, would produce a laser beam that, in a statistical sense, would have a power density close to the diffraction limit. For typical imaging or directed energy applications, this is the figure of merit used to assess the degree of compensation to be achieved. Also, as turbulence is a random process, there will be occasions when higher order aberrations are strong enough to further disrupt the laser beam profile.
Traditional Adaptive Optics (AO) Atmospheric Compensation
One technique that has been used to mitigate the fading problem in free space optical (FSO) laser communication links, is traditional adaptive optics (AO) atmospheric compensation. In principle, an AO compensation system on the transmitter can improve the condition of the beam at the receiver and help to reduce the fluctuation in received power. An adaptive optics compensation system on the receiver should also reduce the size of the spot at the detector plane.
Unfortunately, under conditions like those of the Wave-Optic Propagation Simulations described above, the performance of phase only AO compensation systems is very limited. The reason lies in the difference in the strength and distribution of the turbulence along a horizontal path compared to the vertical path of typical atmospheric AO compensation systems (e.g. astronomical adaptive optics). With turbulence roughly uniformly distributed along the path, phase errors close to the transmitter evolve into strong intensity variations at the far end of the path. Similarly, the light from the beacon on the receiver that is used as the source for the transmitter AO compensation system arrives at the transmitter with strong intensity variations. Portions of the received beam with very low intensity lead to erroneous phase measurements and corrupt the calculated wavefront. Even if the wavefront were measured perfectly and the correct phase compensation was applied to the outgoing laser beam, there would still be intensity variations across the laser beam at the receiver. This is because the transmitted laser beam is typically of uniform intensity while the laser beam that would produce uniform intensity at the receiver has intensity variations that mimic those of the laser beam from the beacon. This can be understood quite simply. Consider a portion of the laser beam from the beacon that is of low intensity. The light in that part of the laser beam has been diverged by the atmosphere to other parts of the beam. When the transmitted uniform beam passes through that same path, the light is converged by the atmosphere to produce a region of high intensity. What is needed to properly compensate the transmitter is full conjugation of the complex electric field of the laser beam from the beacon. While techniques exist for doing this, none have been demonstrated in the field.
At the receiver end, the compensation problem is not as daunting. There are the same problems due to regions of low intensity in the receiver pupil mentioned with regard to the transmit side of an AO compensation system. However, if the phase is correctly measured and compensated, then the laser beam spot at the detector should approach a diffraction limited size. This compensation system cannot do anything to reduce the size of variations in the power coming into the receiver. If, due to those variations, no light enters the receiver's entrance pupil, then no amount of phase compensation will increase the received signal power.
On top of these basic physical limitations of AO compensation systems for horizontal paths, there are also large and costly engineering problems due to the high temporal frequency characteristics of atmospheric turbulence. Measurements over a 16 km horizontal path show significant power in the full aperture tilt at temporal frequencies above 1 kHz. Higher order aberrations probably include even higher temporal frequencies. This requires an AO compensation system with a correction bandwidth of a least 1 kHz. Such response characteristics are beyond the performance of any existing AO systems.
Also, in general, some kind of reference wavefront is required for any type of wavefront measurement system, regardless of whether used as part of an OA subsystem in an FSO laser communication system, or some other kind of instrument or system. In any of the wavefront sensor configurations described above, the concept of the “reference wavefront” may be defined. In a phase-shifting interferometer, that reference wavefront is created by the physical surface of the reference flat or sphere used to form the interferometer. For a Hartmann wavefront sensor as described for example, in Applicant's U.S. Pat. No. 6,631,991, incorporated herein by reference as if set forth fully herein, the reference takes the form of the stored locations of the Hartmann spots for each subaperture when the selected reference wavefront is input. The WFS control loop drives the spots toward these reference spot positions within each subaperture. In an ideal system, these subaperture reference positions would all be zero. In any real system, the null positions differ from zero. Typically, these positions are defined by causing a “perfect” plane wavefront to enter the WFS and recording the observed positions of the spots. This effectively calibrates the WFS reference positions so that the output of each subaperture is very close to zero in each axis when a plane wave is presented to the sensor. For many applications of wavefront sensing, the generation of this reference wavefront and its introduction into the wavefront sensor present significant difficulties.
Perhaps the most common problem to be overcome in creating a reference wavefront is that of non-common path optical aberrations. An example of this is the case is where the wavefront sensor uses one range of wavelengths to make its measurement, while it is desired to measure or correct a beam of a different wavelength. Because of chromatic aberrations in the optics, the wavefront that the sensor measures will differ from the wavefront that is to be corrected. In this case, there is no simple way to introduce the proper reference wavefront that will produce the correct measurement. While prior art laser communications systems provide good examples of this problem, the same problems are nevertheless present in all AO systems. Suppose the WFS is calibrated with a plane wave and the spot reference positions recorded. This defines the null wavefront to be one that is flat at the plane of the WFS input. The control system will work to drive the wavefront at that plane toward the flat condition. Consider now the collection optic and receive fiber assembly. If the fiber is precisely at the infinity focus of the collection optic, a plane wave input will be optimally coupled into the fiber. If the fiber is slightly out of the correct focal plane, or if it is translated so that the plane wave that focuses onto the fiber is at an angle to the plane wave used to calibrate the sensor, then the control system will drive the wavefront to be plane at the WFS, but not optimally couple into the fiber. Further, if the beam splitter between the WFS and the collection optic introduces different aberrations on the transmitted and reflected beams, then the control loop again will do the wrong thing by flattening the wavefront at the WFS.
Various several techniques also have been developed for addressing such “beam-fiber coupling” problems, which arise when the receiver fiber is slightly out of the correct focal plane or if it is translated so that the control system drives the wavefront to be planar at the WFS, but not optimally coupled into the fiber. For example, one compensation technique involves careful mechanical adjustment of the angle between the WFS reference and the boresight of the fiber so as to remove this error. Another compensation technique involves calibrating and applying aberrations of the beam splitter to the WFS reference so that the null of the control loop produces the correct wavefront at the collection optic, rather than at the WFS entrance. However, both of these compensation techniques have the same shortcoming: the magnitude of the misalignment between the tracking sensor and the receiver is not fixed; it can vary with temperature, gravity loading, vibration, etc. Over the course of time, the alignment will degrade and the performance of the receiver system will be compromised. Optical aberrations are also dependent on mechanical alignment and will also vary with time.
Also, in a typical Hartmann wavefront sensor, the Hartmann spot positions are determined by using an algorithm such as “centroiding”. To avoid confusion with adjacent spots, it is necessary to extract from the image, a subarray that contains just one spot. This requirement limits the maximum allowable wavefront gradient in each subaperture of the Hartmann sensor. Furthermore, the gradient cannot be so large that the spot moves out of its subarray. This often leads to a limitation in overall system performance.
In summary, optical communications technology has rapidly advanced and is at the point where communication systems via fiber at 40 Gbs are commercially available. There are applications that prevent the use of a fiber connection, such as those involving moving platforms.
However, the transition of the light from fiber to free space presents problems. To achieve very high data rates requires the use of single mode fiber to prevent mode dispersion from corrupting the data stream. Conversion of the light signals to electrical signals at very high rates places constraints on the physical size of the electro-optical sensor. The capacitance of a detector is proportional to its area and, because of its inherent RC time constant, larger detectors are necessarily slower. For these reasons, to couple free space propagating light into a fiber or onto a detector requires that the light be concentrated into a very small region.
In the case of a single mode fiber, the mode diameter is similar to the diffraction limited spot size defined by the numerical aperture (NA) of the fiber. For efficient coupling, the NA of the coupling lens must match the NA of the fiber. This means that the focal spot needs to be close to diffraction limited to properly couple the light from free space to the fiber.
Unfortunately, propagation of light through the atmosphere introduces aberrations in the optical wavefront that prevent reaching the diffraction limit. To some extent, these aberrations can be corrected with adaptive optics. However, under conditions of strong turbulence (such as encountered in a free space link close to the ground), the disturbance of the optical beam include both phase and intensity components. Traditional phase only adaptive optics cannot compensate for these intensity disturbances. The result is that FSO links under these conditions experience deep signal fade events that lead to unacceptable interruptions in the communications link.
Therefore, there is a great need in the art for an improved way of and a means for producing the highest possible resolution at the focal plane or maximizing the power into a communications fiber, and eliminating the physical misalignment between the communications fiber and the wavefront sensor, so as to improve communication link performance, while avoiding the shortcomings and drawbacks of prior art methodologies and apparatus.